Domenico Guastella & Amelia Gangemi Università di Messina
[email protected] [email protected]
Recent research shows that in reasoning tasks, subjects usually produce an initial intuitive answer, accompanied by a metacognitive experience, which has been called Feeling of Rightness (FOR; Prowse Turner & Thompson, 2009). This paper is aimed at further exploring the complimentary experience of Feeling of Error (FOE; Gangemi, Bourgeois-Gironde & Mancini, in press), that is, the spontaneous, subtle sensation of cognitive uneasiness arising from conflict detection during thinking. According to the Metacognitive Dual- Processes Theory (Thompson et al., 2011), the FOE (which exists along a continuum) is an affective response, resulting from the conflict between the fast and automatic response, due to a “heuristic” reasoning process (Type 1 processes), and the normative response. With FOE subjects become conscious that something is wrong in their performance. They may not know what exactly, they may not precisely refer back to a specific moment in their recent performance, but they sense that they did not perform correctly (see, Piattelli- Palmarini, 1994) and this likely produce the activation of the analytical second system of cognitive processing (Type 2 processes) (see also Thompson et al., 2013). FOE can be compared to similar but positive metacognitive phenomena in the domain of perception and memory. For example, when recall is used to test episodic memory, people sometimes fail to retrieve previously encoded information, but express a Feeling of Knowing (FOK) (e.g., Koriat, 2000; Efklides, 2006), that they could recognize the information on a later test (e.g. Nelson & Narens, 1980). Like FOE, other knowledge states appear to be based on similar but negative metacognitive appraisals (i.e. feelings of wrongness, Thompson & Morsanyi, 2012). In the present study, we want to further explore and confirm the FOE phenomenon, which is assumed to accompany the intuitive and erroneous answers usually evoked by those cognitive tasks that typically provoke biases or illusions. Our goal is to reach a deeper understanding of its nature and its possible role in promoting reflection upon and correction of the erroneous response. Indeed, FOE might play a role in mediating the extent and quality of Type 2 thinking. However, in this paper, we want to verify whether FOE is actually present while performing certain reasoning tasks that generate systematic errors, and specifically whether it is reliable, that is does it arise when people are actually mistaken? To this aim, in our study, we use two different experimental tasks known to generate a large amount of impulsive errors: the bat-and-ball problem and the lilypad problem (Frederick, 2005). Both task are examples of the so-called attribute substitution. Individuals confronted with difficult questions often intuitively answer an easier one instead (e.g. De Neys, Rossi & Houdé, 2013), usually without being aware of the substitution. Although reasoners do not deliberately reflect upon their response, and do not know what the correct answer is, it is possible that they could detect the substitution process, showing a substitution
sensitivity.
Method and participants
A total of 227 undergraduates taking an introductory psychology course at the University of Messina (Italy) participated in return for course credit.Their mean age was 23.5 years (SD= 6 years); 139 (69%) were females. Participants provided written informed consent.
Material and procedure
Participants are randomly assigned to one of two groups. Each group is given only one of the two experimental tasks (bat-and-ball problem: n=99; lilypad problem: n=103), in order to avoid carryover effects from one task to the other. In particular, each participant is given a booklet with written instructions and the bat-and-ball (B&B) or the lilypad (LLP) task. We translated them from English into Italian. In the case of the bat-and-ball problem, culturally common items are used, a stamp and an envelope, as baseball is widely ignored in Italy:
An envelope and a stamp cost €1.10 together. The stamp costs €1 more than the envelope. How much does the envelope cost? The correct answer is 5 cents, but
the task typically elicits incorrect answers, such as 10 cents.
The text of the the lilypad problem is the following: In a lake there is a patch
of lilypads. Every day, the patch doubles in size. If it takes 48 days for the patch to cover the entire lake, how long would it take for the patch to cover half of the lake?This problem also typically elicits incorrect responses, usually
24 instead of 47.
Participants are told to read the problem carefully and to solve it individually. After solving the task, participants are asked to fill in the Feeling of Error- Questionnaire (FOE-Q), assessing the level of cognitive discomfort arising from the just completed task. The FOE-Q begins by presenting the cognitive uneasiness sensation to participants: While I was solving the task, I had the
unpleasant sensation that I was not behaving exactly as I should or the way I would have liked to. Afterwards, respondents are asked to indicate whether or
not they experienced this feeling during the task and the degree of this feeling (from 0: “not at all” to 4: “extremely strong”). Participants are subsequently asked to rate this sensation in five dimensions, by using again a 5-point scale (0 = “not at all”, 4 = “extremely strong”): -How intense was this sensation?-How
unpleasant was this sensation during the task?-How unpleasant is this sensation now?-To what extent was this sensation due to the feeling that you didn’t solve the task as you should?-To what extent did this sensation mean there was something wrong or incomplete in the task? The FOE-Q total score
can range from 0 to 24, with higher scores reflectingstronger FOR.
For both the tasks, each participant’s performance in the task was coded as correct/incorrect.
Results
To examine how well the occurrence ratings for the FOE-Q items held together, we used Cronbach’s alpha. This revealed that the internal consistency for the scale is high in the current sample (a = 0.85). We then evaluated whether FOE was felt more by participants who performed badly in the tasks than by participants who did well. For both the problems, there was a difference in the FOE score between the two groups. For the group who gave an incorrect answer the FOE rating was significantly higher (B&B: M=16.2,
SD=5.2, LLP: M=17, SD=4) than for the group giving a correct answer (B&B: M=9.4, SD=4.1, LLP: M=9.6, SD=3.6), BB: U Mann-Whitney = 390, p < .001, LLP: U Mann-Whitney = 234.5, p < .001.)
Conclusion
Our study was designed to investigate just these feelings of error (FOE) experienced by individuals, when they produce incorrect responses in reasoning tasks explicitly designed to elicit those responses. In general, in line with recent studies (e.g. Gangemi et al., in press) our findings indicate that participants who actually failed in the tasks experienced FOE, as measured through the Sense of Error Questionnaire (FOE-Q), to a greater extent, than those who succeeded in them. This confirms that FOE function as reliable signals when errors are objectively present. Analogously to the FORs (Thompson et al., 2011; Thompson & Johnson, 2014), we could explain this accuracy as determined by a detection of the conflict between the intuitive answer and the normative one (e.g. De Neys et al, 2013; Thompson & Johnson, 2014). The fact that this conflict is affecting their judgments, implies that reasoners implicitly adhere to these normative. Indeed, a characteristic of our substitution problem is that their solution is easily understood by participants, when explained (Frederick, 2005). These experiences thus have a critical role as they should motivate a more analytic approach (e.g. Type 2 thinking). Further studies are thus required to verify whether FOE might play a role in mediating the extent and quality of Type 2 thinking.
Bibliography
De Neys, W., Rossi, S., & Houdé, O. (2013). Bats, balls, and substitution sensitivity: Cognitive misers are no happy fools. Psychonomic Bulletin& Review, 20, 269-273
Efklides, A. (2006). Metacognitive and affect: What can metacognitive experiences tell us about the learning process? Educational Research
Review, 1, 3-14.
Frederick, S. (2005). Cognitive Reflection and Decision Making. Journal of Economic Perspectives, 19, 25-42.
Gangemi, A., Bourgeois-Gironde, S., & Mancini, F. (in press). Feelings of error in reasoning – in search of a phenomenon. Thinking & Reasoning.
Koriat, A., (2000). The Feeling of knowing: Some metatheoretical implications for consciousness and control. Consciousness and Cognition, 9, 149-171.
Nelson, T. O., & Narens, L. (1980). Norms of 300 general-information questions: Accuracy of recall, latency of recall, and feeling-of- knowledge ratings. Journal of Verbal Learning & Verbal Behavior, 19, 338-368.
Piattelli-Palmarini M., (1994). Inevitable Illusions. New York: Wiley.
Prowse Turner, J. A., & Thompson, V. A. (2009). The role of training, alternative models, and logical necessity in determinin g confidence in syllogistic reasoning. Thinking & Reasoning, 15, 69–100.
Matching bias on the selection task: It’s fast and it feels good. Thinking & Reasoning, 19, 431-452.
Thompson, V. A., & Morsanyi, K. (2012). Analytic thinking: Do you feel like it? Mind & Society, 11, 93-105.
Thompson, V., Prowse Turner, J.A., & Pennycook, G. (2011). Intuition, reason, and Metacognition. Cognitive Psychology, 63, 107–140. Thompson, V.A., & Johnson, S.J. (2014). Conflict, metacognition, and